RESEARCH ARTICLE3254

Development 140, 3254-3265 (2013) doi:10.1242/dev.094938© 2013. Published by The Company of Biologists Ltd

INTRODUCTIONPhenotypic variability is typical in human disorders. This variabilitycan be caused by a combination of genetic and environmentalfactors. Although we have gained insight into these separate geneticand environmental influences, the synergistic interactions betweenthe two are largely unknown. We sought to understand theseinteractions by examining the effects of embryonic alcohol exposureon craniofacial development using a zebrafish model.

Fetal alcohol spectrum disorders (FASD) encompass the rangeof defects associated with prenatal exposure to ethanol. By someestimates, FASD affect somewhere from 1% of the North Americanpopulation (Sampson et al., 1997) to 2-5% of the USA and WesternEurope (May et al., 2005; May et al., 2007). Defects and deficitsassociated with FASD are variable and lie along a continuum withthe most severe form represented by fetal alcohol syndrome (FAS),which is clinically diagnosed by growth retardation, deficiencies inbrain growth (reduced head circumference and/or structural brainanomaly) and distinct facial features (microcephaly, short palpebralfissures, thin upper lip and/or smooth philtrum) (Jones and Smith,1973). Other craniofacial defects can arise in FAS, including cleft

palate (Swayze et al., 1997). The variability of the defects found inFAS and FASD are partly attributed to the timing and dose of fetalexposure to ethanol (Ali et al., 2011; Sulik, 2005; Sulik et al., 1986).However, genetic factors are also likely to influence FASD, yetthese influences are poorly understood.

Several lines of evidence suggest a genetic component to FASD(Warren and Li, 2005). Different strains of mice, rats, chickens andzebrafish show differing sensitivity to ethanol exposure (Boehm etal., 1997; Debelak and Smith, 2000; Dlugos and Rabin, 2003;Thomas et al., 1998). Recently, work in mouse has shown that theHedgehog (Hh) pathway gene, Cdon, interacts with ethanol (Hongand Krauss, 2012). The Hh pathway has also recently been shownto interact with ethanol during zebrafish neural development (Zhanget al., 2013). In humans, the most compelling evidence for a geneticcomponent to FASD is twin studies showing greater concordancefor FAS between monozygotic than for dizygotic twins (Streissguthand Dehaene, 1993). In a small South African cohort, resistance toFAS associates with carrying the alcohol dehydrogenase (ADH)B*2 allele, which catabolizes alcohol at a faster rate than other ADHisozymes (Viljoen et al., 2001). Together, these studies provideevidence for gene-environment interactions being involved insusceptibility to FASD.

The zebrafish provides an excellent model system in which tostudy gene-ethanol interactions. Zebrafish eggs are fertilizedexternally, simplifying the analysis of zygotic genes that interactwith ethanol. Furthermore, this external development allows for theprecise timing and dosage of ethanol exposure. Zebrafish embryosdisplay FASD defects when exposed to ethanol and the effects ofethanol timing and dosage have been well studied in zebrafish asfar back as 1910 (Ali et al., 2011; Arenzana et al., 2006; Dlugos andRabin, 2003; Lockwood et al., 2004; Loucks and Carvan, 2004;

1Department of Molecular and Cell and Developmental Biology, Institute for Cellularand Molecular Biology and Institute for Neuroscience, University of Texas, Austin, TX78712, USA. 2Department of Medical and Molecular Genetics, Indiana UniversitySchool of Medicine, Indianapolis, IN 46202, USA.

*Author for correspondence (eberhart@austin.utexas.edu)

This is an Open Access article distributed under the terms of the Creative Commons AttributionLicense (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distributionand reproduction in any medium provided that the original work is properly attributed.

Accepted 4 June 2013

SUMMARYHuman birth defects are highly variable and this phenotypic variability can be influenced by both the environment and genetics.However, the synergistic interactions between these two variables are not well understood. Fetal alcohol spectrum disorders (FASD)is the umbrella term used to describe the wide range of deleterious outcomes following prenatal alcohol exposure. Although FASDare caused by prenatal ethanol exposure, FASD are thought to be genetically modulated, although the genes regulating sensitivityto ethanol teratogenesis are largely unknown. To identify potential ethanol-sensitive genes, we tested five known craniofacialmutants for ethanol sensitivity: cyp26b1, gata3, pdgfra, smad5 and smoothened. We found that only platelet-derived growth factorreceptor alpha (pdgfra) interacted with ethanol during zebrafish craniofacial development. Analysis of the PDGF family in a humanFASD genome-wide dataset links PDGFRA to craniofacial phenotypes in FASD, prompting a mechanistic understanding of thisinteraction. In zebrafish, untreated pdgfra mutants have cleft palate due to defective neural crest cell migration, whereas pdgfraheterozygotes develop normally. Ethanol-exposed pdgfra mutants have profound craniofacial defects that include the loss of thepalatal skeleton and hypoplasia of the pharyngeal skeleton. Furthermore, ethanol treatment revealed latent haploinsufficiency,causing palatal defects in ~62% of pdgfra heterozygotes. Neural crest apoptosis partially underlies these ethanol-induced defects inpdgfra mutants, demonstrating a protective role for Pdgfra. This protective role is mediated by the PI3K/mTOR pathway. Collectively,our results suggest a model where combined genetic and environmental inhibition of PI3K/mTOR signaling leads to variability withinFASD.

KEY WORDS: Zebrafish, Alcohol, Pdgfra, FAS, mTOR

Pdgfra protects against ethanol-induced craniofacial defectsin a zebrafish model of FASDNeil McCarthy1, Leah Wetherill2, C. Ben Lovely1, Mary E. Swartz1, Tatiana M. Foroud2 and Johann K. Eberhart1,*

DEVELO

PMENT

3255RESEARCH ARTICLEProtective role for Pdgfra

Stockard, 1910). Yet we know very little about how genetic factorsinfluence ethanol-induced phenotypes, such as perturbations ofproper craniofacial development.

In all vertebrate species, the majority of the craniofacial skeletonderives from the cranial neural crest (Couly et al., 1993; Evans andNoden, 2006; Gross and Hanken, 2008; Knight and Schilling, 2006;Le Lievre, 1978; Noden, 1978; Yoshida et al., 2008). Neural crestcells are generated between neural and non-neural ectoderm. Neuralcrest cells undergo an epithelial-to-mesenchymal transition, whichallows them to migrate into the periphery to form a wide variety ofcell types (Theveneau and Mayor, 2012). In cranial regions, thesecells will migrate into serially reiterated structures called pharyngealarches (Grevellec and Tucker, 2010). In the zebrafish, this occursbeginning at 10 hours post-fertilization (hpf) and ends around 30hpf (Eberhart et al., 2008). Morphogenetic processes within thepharyngeal arches will eventually shape the craniofacial skeletonby 5 days post-fertilization (dpf). In the zebrafish, the firstpharyngeal arch contributes to jaw structures, including Meckel’scartilage and the palatoquadrate, as well as to palatal elements,including the ethmoid plate and trabeculae (Cubbage and Mabee,1996; Kesteven, 1922; Shah et al., 1990; Swartz et al., 2011). Thesecond arch contributes to the jaw support skeleton, including thehyosymplectic, ceratohyal and interhyal cartilages.

Numerous signaling pathways regulate the development of thesecraniofacial elements, including the platelet-derived growth factor(Pdgf) signaling pathway. Pdgfra is a receptor tyrosine kinase,known to regulate a number of processes, including cellularmigration, proliferation and survival (Tallquist and Kazlauskas,2004; Wu et al., 2008). A key effector of Pdgfra signaling is PI3K,which can regulate numerous cell behaviors, including migrationvia activation of small GTPases and survival and growth viaactivation of mTOR (Downward, 2004; Klinghoffer et al., 2002;Zhou and Huang, 2010). In both mouse and zebrafish, most, if notall, cranial neural-crest cells express Pdgfra and Pdgfra function isrequired in the neural crest (Eberhart et al., 2008; Soriano, 1997;Tallquist and Soriano, 2003). Although the precise function ofPdgfra is not known in mouse, in zebrafish Pdgfra is necessary forproper neural crest cell migration (Eberhart et al., 2008).

The work described here uncovers a second role for pdgfra in theneural crest: protection against ethanol-induced teratogenesis.Testing five craniofacial mutants, cyp26b1, gata3, pdgfra, smad5and smoothened, for dominant enhancement of ethanolteratogenesis, we found that only pdgfra heterozygotes and mutantsshowed enhanced craniofacial defects after ethanol exposure. In asmall human dataset with variable prenatal alcohol exposure, wefind support for a gene-ethanol interaction with single nucleotidepolymorphisms (SNPs) in PDGF family members that associatewith different craniofacial phenotypes. In zebrafish, thesusceptibility to craniofacial defects is at least partly due to neuralcrest cell apoptosis following ethanol exposure. The pdgfra-ethanolinteraction is due to the combined genetic and environmentalattenuation at or downstream of the mechanistic target of rapamycin(mTOR), part of the phosphoinositide 3 kinase (PI3K) pathway.Collectively, our data show that genetic screens using zebrafish willhave important implications in FASD research, both in uncoveringgenetic susceptibility loci and in characterizing the mechanismsunderlying gene-ethanol interactions.

MATERIALS AND METHODSDanio rerio (zebrafish) care and useAll embryos were raised and cared for using established protocols(Westerfield, 1993) with IACUC approval from the University of Texas at

Austin. Tg(fli1:EGFP)y1 transgenic embryos (Lawson and Weinstein,2002) are called fli1:EGFP throughout the text. Embryos were treated inembryo media with 1.0% and 0.5% ethanol, 1.5 μM wortmannin (W1628,Sigma), 3 μM rapamycin (S1120, Selleck) and 25 μM caspase inhibitor 3(Cat#264155, Calbiochem). L-leucine has been used at 100 mM in zebrafish(Payne et al., 2012), but we found that a 50 mM solution was sufficient toprovide rescue in our experiments. The pdgfrab1059 (Eberhart et al., 2008),smad5b1100 (Sheehan-Rooney et al., 2013) and smob577 (Varga et al., 2001)alleles have been described previously. The gata3b1075 and cyp26b1b1024

alleles were recovered from a forward genetic screen at the University ofOregon. To determine ethanol tissue concentration relative to mediaexposure, headspace gas chromatography (GC) was used (C.B.L., RuebenA. Gonzales and J.K.E., unpublished).

Morpholino and RNA injectionApproximately 3 nl of morpholinos (Gene Tools), working concentrations3.5 mg/ml ptena/ptenb (Croushore et al., 2005), and 1.2 mM mir140 (Eberhartet al., 2008), were injected into one- or two-cell stage zebrafish embryos.pdgfra mRNA was made as described previously (Eberhart et al., 2008).

ImmunohistochemistryEmbryos were fixed and prepared for immunohistochemistry following theprocedure outlined previously (Maves et al., 2002) using anti-active caspase3 (Promega) and phospho-Histone H3 (Sigma) primary antibodies withAlexa Fluor 568 anti-rabbit IgG (Invitrogen) secondary antibodies. Forquantification of the ratio of cell death, embryos were counterstained withToPro (Invitrogen) and all nuclei within a single representative z plane werealigned through the lateral 1st and 2nd arches, such that the oral ectodermand 1st pouch, but not the mesodermal cores, were visible.

ImmunoblottingMutants were visually sorted from wild-type/heterozygotes by their neuralcrest cell migration defect (Eberhart et al., 2008). Embryos were euthanized,dechorionated and then de-yolked using a fire-pulled glass pipette. Headswere separated from the tail at the posterior end of the otic vesicle.Immunoblotting was performed as described (Tittle et al., 2011) usingprimary antibodies AKT, pAKT, eIF4B, peIF4B and b-actin (all from CellSignaling), and anti-rabbit secondary antibody (Cell Signaling). Membraneswere reprobed using stripping buffer containing 0.2 M glycine, 0.05%Tween-20 (pH 2.5) for 30 minutes at 70°C. Membranes were washed twicewith TBST, reblocked for 1 hour at room temperature and placed in primaryantibody. ImageJ was used to quantify blots (Schneider et al., 2012).

Cartilage and bone staining and area measurementsFive day postfertilization (dpf) zebrafish embryos were stained with AlcianBlue and Alizarin Red (Walker and Kimmel, 2007), then flat mounted(Kimmel et al., 1998). Images were taken with a Zeiss Axio Imager-AImicroscope, and palate and neurocranial lengths and widths and jaw andjaw-support elements were measured using AxiovisionLE software(AxioVs40 V4.7.1.0). All graphs were made in Microsoft Excel 2011. Weused ANOVA followed by a Tukey-Kramer post-hoc test for all statisticalanalyses.

Confocal microscopy and figure processingConfocal z-stacks were collected on a Zeiss LSM 710 using Zen software.Images were processed in Adobe Photoshop CS.

Statistical analysisOrigin 7.0 was used for one-way ANOVA and Tukey’s range test for thecell death, cell death ratio, palate and neurocranial, and pharyngeal archmeasurement analysis, statistical significance was set at 0.05. GraphpadPrism 5.02 was used for the gas chromatography data.

Human samplesThe published data were collected as part of an ongoing internationalconsortium, the Collaborative Initiative on Fetal Alcohol SpectrumDisorders (CIFASD). Participants were recruited from three sites (SanDiego and Los Angeles, CA, and Atlanta, GA, USA). This study wasapproved by the Institutional Review Board at each site. All participantsand/or their parent(s)/legal guardian(s) provided written informed consent. D

EVELO

PMENT

3256 RESEARCH ARTICLE Development 140 (15)

As part of the study visit, each participant was examined by a traineddysmorphologist who completed a standardized, uniform assessment (Joneset al., 2006). Individuals with a recognizable craniofacial syndrome otherthan FAS were excluded. An objective classification system, based solelyon structural features (palpebral fissure, philtrum and vermillion border)and growth deficiency (head size and height and/or weight) consistent withthe revised Institute of Medicine criteria (Hoyme et al., 2005), was used toclassify subjects. Under this scheme, a participant could receive apreliminary diagnosis of FAS, no FAS, or deferred (Jones et al., 2006).Alcohol exposure data were collected at the interview or from a review ofavailable study data. The extent of reported prenatal alcohol exposureinformation was then classified into one of three categories: none, minimal(>1 drink/week average and never more than two drinks on any oneoccasion during pregnancy) and greater than minimal (>4 drinks/occasionat least once/week or >3 drinks/week). Alcohol exposure was confirmedvia review of records or maternal report, if available (Mattson et al., 2010).

As part of the study visit, a three-dimensional facial image was collectedusing the 3dMD system (3dMD, Atlanta, Georgia). Landmarks wereidentified on the 3D model and used to obtain linear measurements.Replication of landmark placement was required, with less than 2 mmdifference per linear measurement. If the tolerance criterion was not met, athird measurement was taken and the average of the two closestmeasurements was chosen for analysis. For bilateral measurements, onlythe left side was used in analyses. Although a set of 16 standardanthropometric measurements (Moore et al., 2007) were obtained, for thisstudy we only used a subset: inner canthal width, outer canthal width, lowerfacial height, lower facial depth and midfacial depth – which best modeledthe craniofacial defects seen in the zebrafish model.

Genomewide single nucleotide polymorphism (SNP) genotyping wascompleted at the Johns Hopkins GRCF SNP Center using OmniExpressgenome array, which includes over 700,000 SNPs. Standard review of bothsample and SNP was performed. Race and ethnicity were reported by theparticipant or the parent/guardian as part of the study visit and were thenconfirmed based on the SNP genotypes using a principal componentanalysis. To reduce phenotypic heterogeneity due to race, we limited theanalyses to those subjects who, based on their SNP genotypes, were ofEuropean-American descent (n=102).

The regulation and function of PDGF signaling is highly conservedacross vertebrate species; therefore, we identified the SNPs within each ofthe five genes in the PDGF pathway. Owing to the small size of PDGFA(12,700 bp), no SNPs were genotyped or analyzed in this gene; therefore,we could only test the role of five out of the six genes in the PDFG pathway.Univariate analysis was performed for each SNP using linear regressionframework to test whether the SNP genotype × alcohol exposure interactionaccounted for a significant proportion of the variation in the five keyanthropometric measures. Within the model, we also included the maineffects of SNP genotype and alcohol exposure. All measures were correctedfor age and gender. We used a Bonferroni correction for the number of genes(five genes) being tested, which resulted in a significance threshold of 0.01(0.05/5).

Ethics statementInstitutional Review Board approval was obtained at each site. Allparticipants and/or their parent(s)/legal guardian(s) provided writteninformed consent. All zebrafish embryos were raised and cared for usingestablished protocols with IACUC approval (Westerfield, 1993).

Human GWASThe human dataset is from an ongoing international consortium, theCollaborative Initiative on Fetal Alcohol Spectrum Disorders, and has beendescribed previously (Moore et al., 2007) (supplementary material TableS2). Genotyping was performed at the Center for Inherited DiseaseResearch (CIDR). All DNA sources consisted of saliva samples. A total of731,442 SNPs were genotyped on the OmniExpress array, of which 728,140SNPs passed CIDR genotyping control and were released for analysis. Outof the 240 samples submitted for genotyping, three samples were droppedby CIDR due to poor quality genotyping. Two additional samples weredropped due to chromosomal abnormalities. All remaining individuals were

genotyped on at least 98% of the SNPs. All samples were examined forcryptic relatedness and population stratification. There were 16 half-siblingpairs and 37 full sibling pairs in the sample. A principal component-basedanalysis was performed in eigenstrat (Price et al., 2006) on both the sampledata and HapMap reference samples to assign subjects to racial groups. Thefinal European American sample consisted of 102 individuals, of which 102had both a 3D image and alcohol exposure information.

SNPs were included for analysis if the call rate was greater than 98% inthe entire sample, and SNPs were removed if the minor allele frequencywas less than 0.01 or if there was significant deviation from Hardy-Weinberg equilibrium (P<10−6). The final dataset consisted of 688,359, ofwhich 118 were analyzed in the PDGF pathway.

All SNPs analyzed are in supplementary material Table S2.

RESULTSpdgfra and ethanol synergistically interactWe sought to identify gene-ethanol interactions that might underliethe craniofacial variability observed in FASD. We reasoned thatsome genetic loci that influence ethanol teratogenicity would causecraniofacial phenotypes when mutated. We initially screened fivecraniofacial mutants for ethanol sensitivity: smob577, cyp26b1b1024,gata3b1075, smad5b1100 and pdgfrab1059. For this initial screen,embryos were placed in media containing 1% ethanol from 6 hourspost-fertilization (hpf) until 5 dpf, when craniofacial phenotypeswere examined. At this concentration, the tissue levels of ethanol are~41 mM (188.6 mg/dl or 0.18 BAC), as determined by headspacegas chromatography, and wild-type embryos are largely unaffected.This is a concentration of ethanol achieved in binge drinkers (Langeand Voas, 2000). Ethanol appeared to interact only with thehypomorphic pdgfrab1059 allele, dramatically exacerbating mutantphenotypes and uncovering latent haploinsufficiency (Figs 1 and 2,discussed in greater detail below), suggesting a synergisticinteraction. We partially rescued these ethanol-induced defects bypdgfra mRNA injection or by morpholino knockdown of miR140,which negatively regulates pdgfra (Fig. 3; supplementary materialFig. S1) (Eberhart et al., 2008), demonstrating specificity of theinteraction. The variability that we observed in these rescues wasconsistent with the variation that we previously observed in rescuinguntreated pdgfra mutants and our finding that the craniofacialskeleton is highly sensitive to the overall levels of Pdgfra (Eberhartet al., 2008). There are also likely to be between embryo differencesin signal strength downstream of the receptor, such as altered Ptenfunction, which would impact negative feedback into the signalingpathway (Carracedo and Pandolfi, 2008) to add to the inherentvariability in these rescue experiments. Collectively, these resultsshow that ethanol interacts with a subset of gene products involvedin craniofacial development.

Untreated pdgfra mutants have a loss of the ethmoid plate, whichleads to a significant reduction in the length of the palate (Fig. 1C,asterisk; supplementary material Fig. S2) and have slightly smallerpharyngeal skeletal elements (Fig. 2C,G-I; supplementary materialTable S1). Untreated heterozygous siblings develop normally(Fig. 1B; Fig. 2B,G-I; supplementary material Fig. S2; Table S1).Treatment with 1% ethanol from 10 hpf to 5 dpf had a minor effecton the neurocranium of wild-type embryos, with 11% of treated wildtypes having a reduced ethmoid plate (Fig. 1D,G) and the overalllength of the palate was not significantly reduced (supplementarymaterial Fig. S2). Treated wild-type embryos displayed a slightlysmaller hyosymplectic (Fig. 2I; supplementary material Table S1)cartilage; however, Meckel’s cartilage and the palatoquadrateappeared unaffected (Fig. 2G,H; supplementary material Table S1).Statistical analyses of the size of these elements showed that, indeed,only the hyosymplectic was reduced by ethanol. D

EVELO

PMENT

Ethanol-treatment caused craniofacial defects in 62% of pdgfraheterozygotes (Fig. 1E,G), a dramatic increase compared with 11%of wild-type embryos with an identical treatment. Becauseheterozygous embryos rarely have craniofacial defects, thissubstantial increase in the number of embryos with craniofacialdefects is specific to the pdgfra/ethanol interaction. The defects

3257RESEARCH ARTICLEProtective role for Pdgfra

observed included gaps in the ethmoid-plate (asterisk and arrow inFig. 1E) and breaks in the trabeculae (arrowheads in Fig. 1E).Ethanol-treated heterozygotes also displayed significantly shorterpalates when compared with untreated heterozygotes and ethanol-treated wild types (supplementary material Fig. S2A).Heterozygosity for pdgfra alone had no effect on the size of thepalatal skeleton and ethanol caused a slight, nonsignificant,reduction to the palate in wild-type siblings (supplementary materialFig. S2A). The palatal skeleton was significantly reduced in ethanol-treated pdgfra heterozygotes, compared with these two comparisongroups. Therefore, ethanol has a specific synergistic effect on palataldevelopment in pdgfra heterozygotes. Furthermore, ethanol causeda statistically significant reduction in the size of the hyosymplecticcartilage in heterozygotes compared with ethanol-treated wild typesand untreated heterozygotes (Fig. 2E,I; supplementary materialTable S1). Because the hyosymplectic is reduced in ethanol treatedwild-type embryos and untreated heterozygotes, this effect appearsadditive. These results show that pdgfra and ethanol interactsynergistically in palatal development and that the hyosymplectic isfurther reduced in ethanol-treated pdgfra heterozygotes whencompared with treated wild-type embryos.

In 100% of treated pdgfra mutants, compared with ~10% ofuntreated mutants, the anterior neurocranium was completely lost(Fig. 1F). Not surprisingly, this loss led to a statistically significantdecrease in the length of the palate, relative to treated wild-typeembryos and untreated mutants (supplementary material Fig. S2).The ethanol-treated mutant embryos also had a statisticallysignificant reduction of the palatoquadrate, Meckel’s cartilage andthe hyosymplectic compared with untreated mutant and ethanol-treated wild types and heterozygotes (Fig. 2F,G-I; supplementarymaterial Table S1). All of the reductions observed in the ethanol-treated mutants are far beyond that predicted by an additive effectof ethanol-treatment in wild types and loss of pdgfra withoutethanol. Ectodermal and cardiac edema was also present in 100% oftreated pdgfra mutants (not shown). Collectively, these resultsprovide strong support that pdgfra and ethanol synergisticallyinteract during craniofacial development.

The severity and variability of ethanol-induced defects is partlydependent on concentration and developmental timing (Ali et al.,2011; Sulik, 2005). Therefore, we tested these variables acrossethanol-treated pdgfra genotypes. We initially tested 24-hour timewindows for sensitivity and then narrowed down the sensitivewindow. Although treatments initiating after 24 hpf had no effect,we found that a 1% ethanol treatment from 10-24 hpf is the shortesttime window sufficient to fully recapitulate the mutant phenotypesobtained by treatment from 10 hpf to 5 dpf (Fig. 1G; Fig. 2G-I;supplementary material Fig. S2; Table S1). In heterozgyotes, thisshorter exposure still lead to significant differences in theneurocranium, in fact the difference in length between the treatedwild-type and heterozygous embryos was even more marked(Fig. 1G; supplementary material Fig. S2). No significant alterationsto the hyosymplectic were observed (Fig. 2G-I; supplementarymaterial Table S1). These results suggest that the palatal skeleton ismost sensitive to brief ethanol exposure.

We tested lower ethanol concentrations to determine howsensitive pdgfra mutants and heterozygotes were to ethanolteratogenesis. Treatment with 0.5% ethanol, a tissue concentrationof 19 mM (88 mg/dl), from 10 hpf to 5 dpf caused neurocranialdefects in 3% and 13% of wild types and heterozygotes, respectively(Fig. 1G). Under these conditions, 38% of pdgfra mutantscompletely lacked the palatal skeleton (Fig. 1G). This clear increasein the percentage of heterozygous and mutant embryos with palatal

Fig. 1. Ethanol exacerbates pdgfra mutant neurocranial defects andreveals haploinsufficiency. (A-F) Flat-mounted 5 dpf zebrafishneurocrania. Anterior is towards the left. (A,B) Untreated wild types andpdgfra heterozygotes develop normally. (C) Untreated pdgfra mutantshave clefting of the ethmoid plate (asterisk), although the trabeculae aretypically present. (D,E) Neurocrania of embryos treated with 1.0% ethanolfrom 10 hours post-fertilization (hpf ) to 5 days post-fertilization (dpf ). (D)Wild-type embryos are predominantly normal following ethanoltreatment. (E) Ethanol-treated pdgfra heterozygotes display variablepalatal defects, including partial clefting of the ethmoid plate (asterisk),holes in the ethmoid plate (arrow) and breaks in the trabeculae(arrowheads). (F) Ethanol-exposed mutants have an invariant andcomplete loss of the palatal skeleton. (G) Quantification of palatal defectsacross genotypes and treatments. ep, ethmoid plate; tr, trabeculae.

DEVELO

PMENT

3258

defects is associated with reductions in the size of the crest-derivedcartilages (Fig. 2G-I; supplementary material Table S1), althoughthese differences did not reach a level of significance. Shortertreatments with 0.5% ethanol had no apparent effect (data notshown). Thus, attenuation of Pdgf signaling sensitizes embryos toethanol-induced craniofacial defects, and this sensitivity occursroughly when neural crest cells are migrating to and populating thepharyngeal arches, although migration does continue after 24 hpf inzebrafish palatal development (Dougherty et al., 2013).

We analyzed non-crest-derived cartilages to determine whetherethanol had a general effect on cartilage development. First, weexamined the posterior neurocranium because it is of presumedmesoderm origin. Surprisingly, we discovered a previouslyundescribed function of Pdgfra in posterior neurocranialdevelopment, with the size of the posterior neurocranium significantlyreduced in untreated mutants, relative to wild-type embryos(supplementary material Fig. S2). This requirement may relate to the

RESEARCH ARTICLE Development 140 (15)

requirement for Pdgfra function for migration and survival of earlymesoderm precursors (Van Stry et al., 2005; Van Stry et al., 2004). Asthen would be expected, ethanol did interact with pdgfra in thedevelopment of these cartilages (supplementary material Fig. S2) togenerate an overall effect of microcephaly. Second, we analyzedcartilage of the pectoral fin and found no difference between ethanol-treated mutants, ethanol-treated wild-type embryos and untreatedmutants (supplementary material Fig. S3). This result suggests thatthere is no generalized chondrogenic defect in the ethanol-treatedpdgfra mutants. Consistent with this, we found that colIIa wasexpressed appropriately in the posterior neurocranium of ethanol-treated mutants, although, as expected from the phenotype, noethmoid plate or trabeculae condensations were present(supplementary material Fig. S4). Collectively, these findings showthat proper development of the entire neurocranium requires Pdgfrafunction and that ethanol interacts synergistically, with pdgfra havingthe most profound effects on the neural crest-derived skeleton, onwhich we focus.

The regulation and function of Pdgf signaling is highly conservedacross vertebrate species, prompting us to seek evidence from apublished human sample (Jones et al., 2006) for interactionsbetween PDGF family members and prenatal ethanol exposure. Weasked whether there was evidence for ethanol use/SNP interactionon human craniofacial features. We tested members of the humanPDGF family, including PDGFB, PDGFC, PDGFD, PDGFRA andPDGFRB, using a well-characterized sample from the USA (seesupplementary material Table S2 for a complete list of the SNPsanalyzed) (Mattson et al., 2010; Moore et al., 2007). Owing to thesmall size of PDGFA (12,700 bp), no SNPs were genotyped oranalyzed in this gene. The best evidence for gene-ethanolinteractions in humans was found in the two PDGF receptor genes.The most significant evidence for a gene-ethanol interaction wasobserved with two SNPs in PDGFRB, when analyzing midfacialdepth (rs2304061, P=3.7×10−6; rs1075846, P<3.5×10−5). TheseSNPs are in modest linkage disequilibrium (r2=0.57) and were alsomoderately associated with lower facial depth (both P<9.9×10−4). Inaddition, a single SNP in PDGFRA was associated with themeasures of outer canthal width and midfacial depth (P=1.3×10−4

and P=1.7×10−4). Together, these results provide evidence thatzebrafish gene-ethanol interactions may be conserved in humans.

Fig. 2. Ethanol induces pharyngealhypoplasia in pdgfra mutants. (A-C) Flat-mounted jaw and jaw support elements,anterior towards the left. Untreated (A) wild-type, (B) pdgfra+/− and (C) pdgfra−/− embryoshave normally shaped cartilages. (D,E) Treatment with 1.0% ethanol from 10 hpfto 5 dpf does not affect the overall shape ofcartilages in (D) wild-type or (E) pdgfraheterozygous embryos. (F) Ethanol treatmentgreatly disrupts the shape of thepalatoquadrate and hyosymplectic cartilages inpdgfra mutants. (G-I) Quantification of theaverage area, in thousands of μm2 of (G)Meckel’s (mc), (H) palatoquadrate (pq) and (I)hyosymplectic (hs) cartilages (standard errorbars are depicted at 1.5 s.e.m., significance isindicated by non-overlapping bars; Moses,1987). Wild type, light bar; pdgfra heterozygote,light-gray bar; pdgfra mutants, dark-gray bar. ch,ceratohyal. Group numbers are the same as forFig. 1.

Fig. 3. pdgfra mRNA partially rescues the ethanol-induced defects.(A-C) pdgfra mutant 5 dpf neurocrania. (A�-C�) Corresponding jaw andjaw support elements. Anterior is towards the left. (A,A�) Untreated pdgframutants typically have trabeculae (n=13/14). (B,B�) Ethanol treatmentfrom 10-24 hpf causes trabeculae loss in all pdgfra mutants (n=10/10).(C,C�) pdgfra mRNA injection partially rescues the trabeculae (C,arrowhead, n=3/6), and the jaw and jaw support defects in 1.0% ethanol-treated pdgfra mutants (note the shape of the cartilage elements in C�). D

EVELO

PMENT

Neural crest cell apoptosis increases in ethanol-exposed pdgfra mutantsHigh doses of ethanol can cause neural crest cell apoptosis(Cartwright and Smith, 1995; Sulik, 2005) and Pdgf signaling hasbeen implicated in mesoderm-cell survival (Van Stry et al., 2005).We postulated that the craniofacial defects in ethanol-treatedmutants and heterozygotes could be due, at least in part, to anincrease in cranial neural crest cell death. In support of this model,treating pdgfra mutants with ethanol and caspase-inhibitor IIIpartially rescued the ethanol-induced craniofacial defects (Fig. 4G,compare with Fig. 1C,F). We do not expect a full rescue with thistreatment, because clefting of the ethmoid plate is due to a migratorydefect (Eberhart et al., 2008). Thus, this partial rescue supports themodel in which elevated apoptosis underlies some of the effect ofethanol on pdgfra mutants.

We directly determined the levels of apoptosis using anti-activecaspase immunohistochemistry in the neural crest cell labelingfli1:EGFP transgenic line. We quantified neural crest cell death inthe first and second arch at 20 and 24 hpf, encompassing thedevelopmental time window in which pdgfra mutants are mostsensitive to ethanol (Fig. 4A-D�). We did not observe any increasein cell death in ethanol-treated groups at 20 hpf compared withuntreated controls (not shown). At 24 hpf, there were low levels ofneural crest cell death in all untreated genotypes, ranging from oneto two cells per side of the embryo (Fig. 4E). The frequency of celldeath was slightly increased in ethanol-treated wild-type embryos(3.4 cells/side) and was further increased in ethanol-treatedheterozygotes (4.79 cells/side, Fig. 4E), although this effect was notclearly synergistic. In ethanol-treated pdgfra mutants, we found astatistically significant and non-additive increase in apoptosiscompared with all other groups (Fig. 4E, 10.3 cells/side, Tukey test*P≤0.01). In all ethanol-treated genotypes, there were frequentlynon-fli1:GFP-positive cells undergoing apoptosis just dorsal to the

3259RESEARCH ARTICLEProtective role for Pdgfra

first arch, in the relative position of the trigeminal ganglion. Therewere, however, no global elevations in apoptosis at 24 hpf betweenethanol-treated mutants and sibling counterparts compared withtheir untreated mutants and siblings (supplementary materialFig. S5), suggesting that the pdgfra-ethanol interaction leads to anincrease in neural-crest specific cell death. Quantification of theratio of cell death compared with total cells showed a linear increasein all ethanol-treated groups at 24 hpf (Fig. 4F). Ethanol treatmentdid not reduce cell proliferation at 24 hpf in pdgfra mutants relativeto untreated controls; however, we did observe a rise in cellproliferation in wild-type embryos (supplementary material Fig. S6;n=5 for each group). This rise in proliferation may compensate forthe slight increase of cell death at this time.

Although pdgfra mutants are most sensitive to ethanol from 10-24hpf, the effects of ethanol could potentially continue after thetreatment. Therefore, we analyzed cell death after a 6-hour recoveryperiod from ethanol. Strikingly, we observed a significant andsynergistic increase in cell death at 30 hpf, after ethanol was removedat 24 hpf in pdgfra mutants (14.2 cells/side, Tukey test *P≤0.01,Fig. 4E). The trend for elevated cell death in pdgfra heterozygotes,seen at 24 hpf, was abolished with just 6 hours recovery (Fig. 4E,F).This timing of 24-30 hpf correlates with when Pdgfra ligands beginto be expressed in the epithelia adjacent to the neural crest (Eberhartet al., 2008). These data suggest that pdgfra protects against ethanol-induced neural crest cell death during neural crest condensation inthe arches, and may be a compensatory mechanism against ethanol-induced cell death following ethanol exposure.

Detailed fate maps of the pharyngeal arches in the zebrafish at24 hpf allow us to compare directly the location of cell death to theskeletal structures disrupted across ethanol-treated pdgfragenotypes. At 24 hpf, neural-crest cells condensing on the roof ofthe oral ectoderm contribute to most of the anterior neurocranium(Crump et al., 2006; Eberhart et al., 2006). In addition, cells located

Fig. 4. Ethanol-treated pdgfra mutantsexhibit an increase in neural crest celldeath. (A-D) Confocal images of 24 hpf anti-active caspase 3-stained fli1;EGFP embryos.(A�-D�) Active-caspase 3 staining alone.Anterior is towards the left. (A-C�) Untreatedpdgfra mutants, ethanol-treated wild typesand ethanol-treated heterozygotes have lowlevels of neural crest apoptosis. (D,D�) Ethanol-treated pdgfra mutants have greatly elevatedlevels of neural crest apoptosis. Arrowheadsindicate apoptotic neural crest cells. (E) Quantification of cell death acrossgenotypes at 24 hpf and 30 hpf, after ethanolwas removed at 24 hpf (untreated control,light bars; ethanol, dark bars). Ethanol-treatedpdgfra mutants show significant elevation incell death compared with all other genotypes(one-way ANOVA, *P≤0.05). (F) Quantificationof the ratio of cell death relative to total crestcells (untreated control, light bars; ethanol,dark bars; black bars above cell death ratioindicate significant differences using one-wayANOVA, *P≤0.05). (G) 5 dpf pdgfra mutanttreated with 1.0% ethanol and 25 μM caspaseinhibitor at 10-24 hpf. Arrowheads indicatepartial rescue of the trabeculae (compare Gwith Fig. 1C,F). Group numbers are indicatedabove the bars in each graph.

DEVELO

PMENT

3260

dorsally in the 1st and 2nd pharyngeal arches contribute to thepalatoquadrate and hyosymplectic (Crump et al., 2004; Crump etal., 2006; Eberhart et al., 2006). These are the skeletal elements thatwere reduced to the greatest extent in ethanol-treated pdgframutants, and increased cell death in ethanol-treated mutants andheterozygotes occurred in these precursor regions (supplementarymaterial Fig. S7). The regions of enhanced cell death are alsoadjacent to the oral ectoderm and pharyngeal pouches, sources ofligands for Pdgfra (Eberhart et al., 2008). Collectively, these datashow that the pdgfra-ethanol interaction causes an increase in neuralcrest cell death starting at 24 hpf, which contributes to the ethanol-induced craniofacial defects at 5 dpf.

Combined loss of Pdgf signaling and ethanolexposure impinges on mTOR signalingIn combination with our previous analyses (Eberhart et al., 2008),our results show that Pdgf signaling functions in both neural crestmigration and protection from ethanol-induced apoptosis. A mainpdgfra effector regulating cell survival and migration in vitro and inXenopus mesoderm cells is phosphoinositide 3 kinase (PI3K)(Downward, 2004; Van Stry et al., 2005; Xiong et al., 2010).Additionally, PI3K is the major effector of Pdgfra regulatingcraniofacial development in mouse (Klinghoffer et al., 2002). Indifferent in vitro analyses, ethanol has been suggested todownregulate several components of the PI3K pathway, includingPI3K itself, AKT and mTOR, both downstream of PI3K (Guo etal., 2009; Hong-Brown et al., 2010; Vary et al., 2008; Xu et al.,2003). We hypothesized that PI3K signaling mediates Pdgfraprotection from ethanol-induced apoptosis. We increased PI3Ksignaling by injecting untreated and ethanol-treated hypomorphicpdgfra mutants with morpholinos against the negative regulator ofPI3K, pten (Croushore et al., 2005). These injections left wild-typeand heterozygous embryos largely unaffected when compared withcontrols (supplementary material Figs S8, S9). Upregulating thePI3K pathway in either untreated or ethanol-treated mutants nearlyfully rescued the palate defects when compared with untreated andethanol-treated mutant controls (Fig. 5A-D). Injection of ptenmorpholinos also rescued the pharyngeal skeletal elements ofethanol-treated mutants to a nearly wild-type appearance (Fig. 5A�,compare with 5D�).

We wanted to test whether activation of this pathway at the levelof mTOR could also rescue the ethanol-induced mutant phenotypes.L-Leucine has been shown to increase mTOR activity, leading to arise in cell growth and survival (Hong-Brown et al., 2012; Kimballand Jefferson, 2006). Supplementing ethanol with 50 mM L-leucine

RESEARCH ARTICLE Development 140 (15)

partially rescued the ethanol-treated phenotype (Fig. 5E-F�), andhad no apparent effect on either untreated heterozygote or wild-typesilbings (supplementary material Fig. S10). However, a 10-24 hpfethanol treatment with L-leucine treatment alone from 24 to 48 hpfdid not rescue ethanol-induced defects in mutants (supplementarymaterial Fig. S11, compare with Fig. 1). These data show thatelevating either PI3K or mTOR signaling is sufficient to protectagainst ethanol-induced craniofacial defects in a Pdgfra attenuatedembryo.

If the pdgfra/ethanol interaction is due to inhibition of the PI3Kpathway downstream of Pdgfra, then pharmacological blockade ofthe PI3K pathway should recapitulate the ethanol defects in pdgframutants. Consistent with this model, treating pdgfra mutants withinhibitors of either PI3K (wortmannin) or mTOR (rapamycin) from10 hpf to 24 hpf phenocopied the ethanol-induced defects of thepalatal skeleton (Fig. 6F,I,L,M; supplementary material Fig. S12)and caused reductions in the pharyngeal skeleton compared withmutant controls and inhibitor-treated wild-type and heterozygouscounterparts (Fig. 7F,I,L-O). Furthermore, these treatments alsouncovered latent haploinsufficiency, causing variable defects,including those affecting the ethmoid plate and trabeculae(Fig. 6H,K,M). Although there are subtle differences between thetreatments that could be due to dose or mechanism of action, theoverall pattern of pharyngeal reduction is highly similar acrosstreatments (Fig. 7). These data suggest that attenuation of the PI3K-mTOR pathway may underlie the pdgfra/ethanol interaction.

To directly test the effects of ethanol on the PI3K pathway inpdgfra mutants, we detected phosphorylated forms of the PI3Ktarget AKT and the mTOR target eIF4B (Fig. 8). Heads fromuntreated and ethanol-treated pdgfra mutants and siblings werecollected at 24 hpf. Downstream of mTOR, ethanol-treated mutantsshowed a decrease in levels of phospho-eIF4B compared withuntreated embryos and ethanol-treated controls (Fig. 8C,D).Upstream of mTOR, we observed an increase in levels of phospho-AKT in both ethanol-treated mutants and controls compared withtheir untreated counterparts (Fig. 8A,B). These data are consistentwith a growing body of evidence, primarily in vitro and in cancermodels, that shows that inhibiting mTOR leads to an increase inAKT phosphorylation (Carracedo and Pandolfi, 2008; Hsu et al.,2011; Soares et al., 2013). To validate this feedback in thedeveloping zebrafish, we assayed phospho-AKT levels inrapamycin-treated embryos. Indeed, blocking mTOR led to anincrease in levels of phosphorylated AKT in both mutants and wildtypes compared with DMSO-treated controls (supplementarymaterial Fig. S13), similar to the effects of ethanol. These data

Fig. 5. Elevating PI3K/mTOR signaling rescues the craniofacial defects in ethanol-treated pdgfra mutants. (A-F�) 5 dpf pdgfra mutantneurocrania (A-F) and the corresponding pharyngeal skeletal elements (A�-F�). (A,A�) Untreated mutants have clefting of the ethmoid plate (asterisk,n=10/10). (B,B�) Treatment with 1.0% ethanol at 10-24 hpf causes loss of the palatal skeleton and hypoplasia of the pharyngeal skeleton (n=12/12). (C-D�) pten morpholino (MO) injection rescues the craniofacial phenotypes of both (C,C�, n=4/4) untreated and (D,D�, n=8/20) ethanol exposed pdgframutants. (E,E�) Most L-leucine-treated mutants have trabeculae (n=12/17; five embryos lacked trabeculae). (F,F�) Supplementing ethanol-treatedmutants with L-leucine causes rescue of the trabeculae (n=10/32). Arrowheads mark the partially rescued ethmoid plates. Asterisk in E indicates cleftingof the ethmoid plate. D

EVELO

PMENT

support a model in which ethanol, in combination with loss of Pdgfsignaling, attenuates the activity of mTOR, leading tohypophosphorylation of targets such as eIF4B, but also an increasein phosphorylated AKT.

DISCUSSIONCollectively, our results provide insight into the geneticsusceptibility to ethanol-induced defects. Using zebrafish, we show

3261RESEARCH ARTICLEProtective role for Pdgfra

that attenuated Pdgf signaling and ethanol exposure synergisticallyinteract to cause variable craniofacial defects. This interaction maybe conserved in humans, through our analysis of the human datatsetherein, although further analysis in humans is necessary to confirmthis finding. We suggest a model in which attenuated growth factorsignaling interacts with ethanol to reduce the activity ofPI3K/mTOR signaling, at or downstream of mTOR, thus causingdevelopmental perturbations. At low levels of attenuation, whetherfrom slight reduction in growth factor signaling or from low dosesof ethanol, the system can compensate. However, those embryoswith reduced growth factor signaling, even reductions that wouldnot cause defects alone, that receive a ‘second hit’ via ethanol, candevelop craniofacial defects.

pdgfra regulates both neural-crest cell migrationand protection from ethanol-induced apoptosisPdgf signaling is essential for midfacial development acrossvertebrate species (Eberhart et al., 2008; Soriano, 1997; Tallquist andSoriano, 2003). At least in zebrafish, Pdgfra regulates the appropriatemigration of neural-crest cells that will generate the midfacial skeleton(Eberhart et al., 2008). It is only in the presence of ethanol that wereveal the requirement of Pdgfra signaling in neural-crest cellsurvival. The increased apoptosis in ethanol-treated pdgfra mutantscould be attributed to the fact that the pdgrab1059/b1059 mutant allele isa hypomorph (Eberhart et al., 2008). In this model, the amount ofPdgfra signaling left in the hypomorph is sufficient for cellularsurvival. However, in the presence of ethanol, these signals no longerpromote survival. This model is consistent with the finding that thereis elevated cell death in Pdgfra-null mice (Soriano, 1997). However,in a neural crest conditional Pdgfra mutant, apoptosis did not appearelevated (Tallquist and Soriano, 2003), which may suggest analternate model. The pdgfra-ethanol interaction could be due to broadinhibition of mTOR signaling, across growth factor pathways. Thezebrafish mutation project (http://www.sanger.ac.uk/Projects/D_rerio/zmp/) has identified a putative null pdgfra allele. Analysisof this mutant allele will aid in distinguishing between these twopossibilities.

Correct Pdgfra signaling relies heavily on PI3K activation. Inmouse, knockout of the PI3K domain of Pdgfra leads to craniofacialdefects, which mimic the full knockout phenotype (Klinghoffer etal., 2002). In frog, the PI3K domain of Pdgfra is necessary andsufficient for proper mesoderm migration and survival (Nagel et al.,2004; Van Stry et al., 2005). In our studies, increasing PI3K-mediated signaling through pten morpholino knockdown leads to anear wild-type phenotype in both untreated and ethanol-treatedpdgfra mutants. This finding suggests that increasing PI3K functionin ethanol-treated pdgfra mutant embryos rescues both survival andmigration, as the midline defects are due to a migratory defect.

Although elevating PI3K signaling does rescue the defects inpdgfra mutants, understanding the precise mechanism will requireadditional studies. GTPases, which promote migration, and AKT,which supports survival and migration, rely on the PIP3 substrate ofPI3K (Zhou and Huang, 2010). Thus, it is possible that themigratory and survival functions of Pdgfra are controlled bydifferent signaling components immediately downstream of PI3K.

mTOR most likely regulates cell survival and growth in ethanol-treated embryos. L-leucine supplementation does not rescue themidline defects in our pdgfra mutants, which are caused by amigratory defect. In ethanol-treated mutants, we observe a decreasein the levels of activated eIF4B and an increase in levels of activatedAKT, suggesting that ethanol impairs growth signaling at the levelof mTOR or below. Further evidence that ethanol may be impairing

Fig. 6. Wortmannin and rapamycin phenocopy ethanol-treatedpdgfra mutants. (A-L) 5 dpf zebrafish neurocrania embryos were treatedat 10-24 hpf with (A-C) DMSO, (D-F) 1.0% ethanol, (G-I) 1.5 μMwortmannin or (J-L) 3 μM rapamycin. Asterisks indicate ethmoid-platedefects (ep); arrowheads indicate trabeculae defects (tr). (M) Graph ofpercentage defects found in all three genotypes, across treatments.Unaffected, wild-type phenotype; ep, ethmoid-plate defects (arrow andasterisk in K); tr, trabeculae defects (arrowheads in H).

DEVELO

PMENT

3262

mTOR activity specifically is the finding that phosphorylated levelsof AKT are increased in ethanol treatment. A similar increase inactivated AKT is observed in our model via rapamycin treatment ofpdgfra mutants. Blocking mTOR function may release inhibitoryfeedback loops; for example, pancreatic cancer cells treated withrapamycin show elevated levels of phosphorylated AKT comparedwith untreated controls (Soares et al., 2013). It is interesting tospeculate that the upregulation of AKT in ethanol-treated siblingsmay buffer against deleterious phenotypes by elevating cellproliferation. However, a more extensive analysis of the PI3K andmTOR pathways will be necessary to fully understand how Pdgfra-mediated migration and protection are controlled intracellularly.

RESEARCH ARTICLE Development 140 (15)

Ethanol may broadly disrupt growth-factorsignaling pathwaysThe interaction observed between ethanol and pdgfra supports amodel in which PI3K-dependent growth factor genes interact withethanol. Notable among these genes is insulin receptor (insr). Insrfunctions in proper brain development. Reducing the expression ofInsr in newly born rat pups causes defects similar to those found inrat models of FASD (de la Monte et al., 2011). Increasing insulinreceptor expression in a Drosophila model of FASD rescuesneurobehavioral defects caused by ethanol exposure (McClure etal., 2011). These data, along with our observations of pdgfra,strongly suggest that growth factor genes play a significant role inthe susceptibility to ethanol teratogenesis. Because growth factorsignaling is used throughout development, attenuated growth factorsignaling, along with ethanol exposure, could explain a large part ofthe variability found in FASD. This model is consistent with thehuman dataset in which the phenotypes associated withgene/ethanol interactions are different between PDGFRA andPDGFRB. As the human dataset grows, it will be of great interestto examine further the extent of interactions between ethanol andPI3K-dependent growth factors.

Zebrafish as a model of gene-environmentinteractionsThe zebrafish provides a powerful tool in uncovering both thegenetics and the mechanisms of FASD susceptibility. Owing to thearray of genetic resources available and the ease of administeringethanol to the externally fertilized embryos, it is feasible to screenhundreds of genes rapidly. As we have done here, findings inzebrafish can be used to direct analyses in human datasets that maynot yet be large enough for genome-wide power. Although thehuman dataset we analyzed is small, by focusing on a small numberof genes, the likelihood of detecting false associations was greatlydecreased. After correcting for the testing of SNPs across five genes,and therefore five essentially independent tests, the associations wefound in the genes encoding for the human PDGF receptors weresignificant, with P<0.01. This significant association even

Fig. 7. Wortmannin or rapamycin inducespharyngeal skeleton hypoplasia in pdgfra mutants.(A-L) Embryos were treated at 10-24 hpf with (A-C)DMSO, (D-F) 1.0% ethanol, (G-I) 1.5 μM wortmannin or(J-L) 3 μM rapamycin. (M-O) Graphs depict the averagearea sizes, in thousands of μm2 in (M) the Meckel’s (mc),(N) palatoquadrate (pq) and (O) hyosymplectic (hs)cartilages. D, DMSO; E, ethanol; W, wortmannin, R,rapamycin [standard error bars are depicted at 1.5s.e.m., significance is indicated by non-overlapping bars(Moses, 1987)]. Wild-type, light bar; pdgfraheterozygote, light-gray bar; mutant, dark-gray bar; ch,ceratohyal. Group numbers are the same as for Fig. 6.

Fig. 8. Ethanol alters phosphorylation of AKT and pEIF4B in pdgframutants. Heads of 24 hpf pdgfra mutant (−/−) and pdgfra sibling (+/?)from untreated controls or embryos treated with 1.0% ethanol from 10-24hpf were used. (A) Immunoblot was stripped and reprobed for total AKT,phospho-AKT (pAKT) and β-actin as a loading control. Both ethanol-treated pdgfra siblings and pdgfra mutants have elevated levels ofphospho-AKT compared with untreated counterparts. (B) Ratio of pAKTto total AKT across three immunoblots. (C) Immunoblot was stripped andreprobed for total eIF4B (eIF4B), phospho-eIF4B (peIF4B) and β-actin as aloading control. pdgfra mutants have decreased levels of phospho-eIF4Bcompared with untreated mutants, and both untreated and ethanol-treated siblings samples. (D) Ratio of phospho-eIF4B to total eIF4B acrossthree immunoblots. D

EVELO

PMENT

3263RESEARCH ARTICLEProtective role for Pdgfra

withstands a more strict correction of treating all 118 SNPs asindependent tests (0.05/118=4.2×10−4). Thus, the zebrafish modelguided focused studies in a human dataset and allowed us to detecta gene-ethanol interaction contributing to the variation incraniofacial measures. Importantly, owing to the conservation ofgene function across vertebrate species, we can understand themechanism of these interactions using the zebrafish model system.Collectively, our findings provide a predictive mechanistic modelfor some gene-ethanol interactions that underlie the craniofacialdefects in FASD.

Although we have focused on the craniofacial aspect of FASD,the zebrafish provides an excellent model in which to study theneural aspects of FASD. The brain is the most commonly affectedorgan in FASD, with defects to the cerebellum and oligodendrocytesbeing common (Hoyme et al., 2005; Swayze et al., 1997). Pdgfra isimportant for oligodendrocyte maturation and migration in mouse(McKinnon et al., 2005), although the function of pdgfra inzebrafish oligodendrocyte development has not been tested. Itwould be of interest to determine if oligodendrocytes are disruptedin ethanol-treated pdgfra mutants.

Owing to its genetic tractability the zebrafish has long been avaluable model system for understanding developmental processes.We currently understand an enormous amount regarding thegenetics regulating development from work in numerous modelsystems. Due to the level of this understanding, we are now wellpoised to gain substantial knowledge about how the environmentimpinges on the genetic networks underlying development (socalled Eco-Devo). Our work here and recent work examining gene-environment interactions underlying ototoxicity (Coffin et al., 2010)demonstrate that the same characteristics (e.g. genetic amenability,ease of imaging and external fertilization) that make zebrafish usefulfor understanding development make it useful for understandingEco-Devo.

AcknowledgementsWe thank Jeffrey Gross for help on the western blot analysis and JenniferMorgan, for use of her lab’s analytical software. For the GC analysis, we thankboth Rueben Gonzalez and Regina Nobles. We would also thank MichaelCharness and Ed Riley for their advice and input in the preparation of thismanuscript, and Young-Jun Jeon for his advice regarding ethanol and mTOR.

FundingFunding to support this research was provided by the National Institutes ofHealth/National Institute of Dental and Craniofacial Research (NIH/NIDCR)[R01DE020884]; by grants from the The Foundation for Alcohol Research(ABMRF) to J.K.E. and from the NIH/National Institute on Alcohol Abuse andAlcoholism (NIAAA) [U01AA014809 10 to T.M.F.]; by a Bruce/Jones Fellowshipand a NIH/NIAAA grant [F31AA020731 to N.M.]; and by a NIH/NIAAA grant[F32AA021320 to C.B.L.]. We also thank Charles B. Kimmel and the Universityof Oregon Fish researchers and staff for their contributions to the forwardgenetic screen [P01 HD22486 to C. B. Kimmel]. Deposited in PMC forimmediate release.

Competing interests statementThe authors declare no competing financial interests.

Author contributionsN.M. and J.K.E. conceived, wrote and edited the manuscript. N.M. performedall of the characterizations of the pdgfra/ethanol interaction in zebrafish. L.W.and T.M.F. performed and analyzed the human data and edited themanuscript. M.E.S. assisted with the initial discovery of the pdgfra/ethanolinteraction, performed the miR140 morpholino and in situ hybridizationexperiments and edited the manuscript. C.B.L. determined the tissue levels ofethanol in treated embryos and edited the manuscript.

Supplementary materialSupplementary material available online athttp://dev.biologists.org/lookup/suppl/doi:10.1242/dev.094938/-/DC1

ReferencesAli, S., Champagne, D. L., Alia, A. and Richardson, M. K. (2011). Large-scale

analysis of acute ethanol exposure in zebrafish development: a critical timewindow and resilience. PLoS ONE 6, e20037.

Arenzana, F. J., Carvan, M. J., 3rd, Aijón, J., Sánchez-González, R., Arévalo,R. and Porteros, A. (2006). Teratogenic effects of ethanol exposure onzebrafish visual system development. Neurotoxicol. Teratol. 28, 342-348.

Boehm, S. L., II, Lundahl, K. R., Caldwell, J. and Gilliam, D. M. (1997). Ethanolteratogenesis in the C57BL/6J, DBA/2J, and A/J inbred mouse strains. Alcohol14, 389-395.

Carracedo, A. and Pandolfi, P. P. (2008). The PTEN-PI3K pathway: of feedbacksand cross-talks. Oncogene 27, 5527-5541.

Cartwright, M. M. and Smith, S. M. (1995). Stage-dependent effects of ethanolon cranial neural crest cell development: partial basis for the phenotypicvariations observed in fetal alcohol syndrome. Alcohol. Clin. Exp. Res. 19, 1454-1462.

Coffin, A. B., Ou, H., Owens, K. N., Santos, F., Simon, J. A., Rubel, E. W. andRaible, D. W. (2010). Chemical screening for hair cell loss and protection in thezebrafish lateral line. Zebrafish 7, 3-11.

Couly, G. F., Coltey, P. M. and Le Douarin, N. M. (1993). The triple origin of skullin higher vertebrates: a study in quail-chick chimeras. Development 117, 409-429.

Croushore, J. A., Blasiole, B., Riddle, R. C., Thisse, C., Thisse, B., Canfield, V.A., Robertson, G. P., Cheng, K. C. and Levenson, R. (2005). Ptena and ptenb genes play distinct roles in zebrafish embryogenesis. Dev. Dyn. 234,911-921.

Crump, J. G., Maves, L., Lawson, N. D., Weinstein, B. M. and Kimmel, C. B.(2004). An essential role for Fgfs in endodermal pouch formation influenceslater craniofacial skeletal patterning. Development 131, 5703-5716.

Crump, J. G., Swartz, M. E., Eberhart, J. K. and Kimmel, C. B. (2006). Moz-dependent Hox expression controls segment-specific fate maps of skeletalprecursors in the face. Development 133, 2661-2669.

Cubbage, C. C. and Mabee, P. M. (1996). Development of the cranium andpaired fins in the zebrafish Danio rerio (Ostariophysi, Cyprinidae). J. Morphol.229, 121-160.

de la Monte, S. M., Tong, M., Bowling, N. and Moskal, P. (2011). si-RNAinhibition of brain insulin or insulin-like growth factor receptors causesdevelopmental cerebellar abnormalities: relevance to fetal alcohol spectrumdisorder. Mol. Brain 4, 13.

Debelak, K. A. and Smith, S. M. (2000). Avian genetic background modulatesthe neural crest apoptosis induced by ethanol exposure. Alcohol. Clin. Exp. Res.24, 307-314.

Dlugos, C. A. and Rabin, R. A. (2003). Ethanol effects on three strains ofzebrafish: model system for genetic investigations. Pharmacol. Biochem. Behav.74, 471-480.

Dougherty, M., Kamel, G., Grimaldi, M., Gfrerer, L., Shubinets, V., Ethier, R.,Hickey, G., Cornell, R. A. and Liao, E. C. (2013). Distinct requirements forwnt9a and irf6 in extension and integration mechanisms during zebrafishpalate morphogenesis. Development 140, 76-81.

Downward, J. (2004). PI 3-kinase, Akt and cell survival. Semin. Cell Dev. Biol. 15,177-182.

Eberhart, J. K., Swartz, M. E., Crump, J. G. and Kimmel, C. B. (2006). EarlyHedgehog signaling from neural to oral epithelium organizes anteriorcraniofacial development. Development 133, 1069-1077.

Eberhart, J. K., He, X., Swartz, M. E., Yan, Y.-L., Song, H., Boling, T. C.,Kunerth, A. K., Walker, M. B., Kimmel, C. B. and Postlethwait, J. H. (2008).MicroRNA Mirn140 modulates Pdgf signaling during palatogenesis. Nat. Genet.40, 290-298.

Evans, D. J. and Noden, D. M. (2006). Spatial relations between aviancraniofacial neural crest and paraxial mesoderm cells. Dev. Dyn. 235, 1310-1325.

Grevellec, A. and Tucker, A. S. (2010). The pharyngeal pouches and clefts:Development, evolution, structure and derivatives. Semin. Cell Dev. Biol. 21,325-332.

Gross, J. B. and Hanken, J. (2008). Review of fate-mapping studies ofosteogenic cranial neural crest in vertebrates. Dev. Biol. 317, 389-400.

Guo, R., Zhong, L. and Ren, J. (2009). Overexpression of aldehydedehydrogenase-2 attenuates chronic alcohol exposure-induced apoptosis,change in Akt and Pim signalling in liver. Clin. Exp. Pharmacol. Physiol. 36, 463-468.

Hong, M. and Krauss, R. S. (2012). Cdon mutation and fetal ethanol exposuresynergize to produce midline signaling defects and holoprosencephalyspectrum disorders in mice. PLoS Genet. 8, e1002999.

Hong-Brown, L. Q., Brown, C. R., Kazi, A. A., Huber, D. S., Pruznak, A. M. andLang, C. H. (2010). Alcohol and PRAS40 knockdown decrease mTOR activityand protein synthesis via AMPK signaling and changes in mTORC1 interaction.J. Cell. Biochem. 109, 1172-1184.

Hong-Brown, L. Q., Brown, C. R., Kazi, A. A., Navaratnarajah, M. and Lang, C.H. (2012). Rag GTPases and AMPK/TSC2/Rheb mediate the differential D

EVELO

PMENT

3264 RESEARCH ARTICLE Development 140 (15)

regulation of mTORC1 signaling in response to alcohol and leucine. Am. J.Physiol. 302, C1557-C1565.

Hoyme, H. E., May, P. A., Kalberg, W. O., Kodituwakku, P., Gossage, J. P.,Trujillo, P. M., Buckley, D. G., Miller, J. H., Aragon, A. S., Khaole, N. et al.(2005). A practical clinical approach to diagnosis of fetal alcohol spectrumdisorders: clarification of the 1996 institute of medicine criteria. Pediatrics 115,39-47.

Hsu, P. P., Kang, S. A., Rameseder, J., Zhang, Y., Ottina, K. A., Lim, D.,Peterson, T. R., Choi, Y., Gray, N. S., Yaffe, M. B. et al. (2011). The mTOR-regulated phosphoproteome reveals a mechanism of mTORC1-mediatedinhibition of growth factor signaling. Science 332, 1317-1322.

Jones, K. L. and Smith, D. W. (1973). Recognition of the fetal alcohol syndromein early infancy. Lancet 302, 999-1001.

Jones, K. L., Robinson, L. K., Bakhireva, L. N., Marintcheva, G., Storojev, V.,Strahova, A., Sergeevskaya, S., Budantseva, S., Mattson, S. N., Riley, E. P.et al. (2006). Accuracy of the diagnosis of physical features of fetal alcoholsyndrome by pediatricians after specialized training. Pediatrics 118, e1734-8.

Kesteven, H. L. (1922). A new interpretation of the bones in the palate andupper jaw of fishes: Part I. J. Anat. 56, 307-324.

Kimball, S. R. and Jefferson, L. S. (2006). Signaling pathways and molecularmechanisms through which branched-chain amino acids mediatetranslational control of protein synthesis. J. Nutr. 136 Suppl., 227S-231S.

Kimmel, C. B., Miller, C. T., Kruze, G., Ullmann, B., BreMiller, R. A., Larison, K.D. and Snyder, H. C. (1998). The shaping of pharyngeal cartilages during earlydevelopment of the zebrafish. Dev. Biol. 203, 245-263.

Klinghoffer, R. A., Hamilton, T. G., Hoch, R. and Soriano, P. (2002). An allelicseries at the PDGFalphaR locus indicates unequal contributions of distinctsignaling pathways during development. Dev. Cell 2, 103-113.

Knight, R. D. and Schilling, T. F. (2006). Cranial neural crest and development ofthe head skeleton. Adv. Exp. Med. Biol. 589, 120-133.

Lange, J. E. and Voas, R. B. (2000). Defining binge drinking quantities throughresulting BACs. Annu. Proc. Assoc. Adv. Automot. Med. 44, 389-404.

Lawson, N. D. and Weinstein, B. M. (2002). In vivo imaging of embryonicvascular development using transgenic zebrafish. Dev. Biol. 248, 307-318.

Le Lièvre, C. S. (1978). Participation of neural crest-derived cells in the genesis ofthe skull in birds. J. Embryol. Exp. Morphol. 47, 17-37.

Lockwood, B., Bjerke, S., Kobayashi, K. and Guo, S. (2004). Acute effects ofalcohol on larval zebrafish: a genetic system for large-scale screening.Pharmacol. Biochem. Behav. 77, 647-654.

Loucks, E. and Carvan, M. J., III (2004). Strain-dependent effects ofdevelopmental ethanol exposure in zebrafish. Neurotoxicol. Teratol. 26, 745-755.

Mattson, S. N., Foroud, T., Sowell, E. R., Jones, K. L., Coles, C. D., Fagerlund,A., Autti-Rämö, I., May, P. A., Adnams, C. M., Konovalova, V. et al.; CIFASD(2010). Collaborative initiative on fetal alcohol spectrum disorders:methodology of clinical projects. Alcohol 44, 635-641.

Maves, L., Jackman, W. and Kimmel, C. B. (2002). FGF3 and FGF8 mediate arhombomere 4 signaling activity in the zebrafish hindbrain. Development 129,3825-3837.

May, P. A., Gossage, J. P., Brooke, L. E., Snell, C. L., Marais, A. S., Hendricks, L.S., Croxford, J. A. and Viljoen, D. L. (2005). Maternal risk factors for fetalalcohol syndrome in the Western cape province of South Africa: a population-based study. Am. J. Public Health 95, 1190-1199.

May, P. A., Gossage, J. P., Marais, A. S., Adnams, C. M., Hoyme, H. E., Jones, K.L., Robinson, L. K., Khaole, N. C., Snell, C., Kalberg, W. O. et al. (2007). Theepidemiology of fetal alcohol syndrome and partial FAS in a South Africancommunity. Drug Alcohol Depend. 88, 259-271.

McClure, K. D., French, R. L. and Heberlein, U. (2011). A Drosophila model forfetal alcohol syndrome disorders: role for the insulin pathway. Dis. Model. Mech.4, 335-346.

McKinnon, R. D., Waldron, S. and Kiel, M. E. (2005). PDGF alpha-receptorsignal strength controls an RTK rheostat that integrates phosphoinositol 3�-kinase and phospholipase Cgamma pathways during oligodendrocytematuration. J. Neurosci. 25, 3499-3508.

Moore, E. S., Ward, R. E., Wetherill, L. F., Rogers, J. L., Autti-Rämö, I.,Fagerlund, A., Jacobson, S. W., Robinson, L. K., Hoyme, H. E., Mattson, S.N. et al.; CIFASD (2007). Unique facial features distinguish fetal alcoholsyndrome patients and controls in diverse ethnic populations. Alcohol. Clin.Exp. Res. 31, 1707-1713.

Moses, L. E. (1987). Graphical methods in statistical analysis. Annu. Rev. PublicHealth 8, 309-353.

Nagel, M., Tahinci, E., Symes, K. and Winklbauer, R. (2004). Guidance ofmesoderm cell migration in the Xenopus gastrula requires PDGF signaling.Development 131, 2727-2736.

Noden, D. M. (1978). The control of avian cephalic neural crestcytodifferentiation. I. Skeletal and connective tissues. Dev. Biol. 67, 296-312.

Payne, E. M., Virgilio, M., Narla, A., Sun, H., Levine, M., Paw, B. H., Berliner,N., Look, A. T., Ebert, B. L. and Khanna-Gupta, A. (2012). L-Leucineimproves the anemia and developmental defects associated with Diamond-

Blackfan anemia and del(5q) MDS by activating the mTOR pathway. Blood 120,2214-2224.

Price, A. L., Patterson, N. J., Plenge, R. M., Weinblatt, M. E., Shadick, N. A.and Reich, D. (2006). Principal components analysis corrects for stratificationin genome-wide association studies. Nat. Genet. 38, 904-909.

Sampson, P. D., Streissguth, A. P., Bookstein, F. L., Little, R. E., Clarren, S. K.,Dehaene, P., Hanson, J. W. and Graham, J. M., Jr (1997). Incidence of fetalalcohol syndrome and prevalence of alcohol-related neurodevelopmentaldisorder. Teratology 56, 317-326.

Schneider, C. A., Rasband, W. S. and Eliceiri, K. W. (2012). NIH Image toImageJ: 25 years of image analysis. Nat. Methods 9, 671-675.

Shah, R. M., Donaldson, E. M. and Scudder, G. G. (1990). Toward the origin ofthe secondary palate. A possible homologue in the embryo of fish,Onchorhynchus kisutch, with description of changes in the basementmembrane area. Am. J. Anat. 189, 329-338.

Sheehan-Rooney, K., Swartz, M. E., Lovely, C. B., Dixon, M. J. and Eberhart,J. K. (2013). Bmp and Shh signaling mediate the expression of satb2 in thepharyngeal arches. PLoS ONE 8, e59533.

Soares, H. P., Ni, Y., Kisfalvi, K., Sinnett-Smith, J. and Rozengurt, E. (2013).Different patterns of Akt and ERK feedback activation in response torapamycin, active-site mTOR inhibitors and metformin in pancreatic cancercells. PLoS ONE 8, e57289.

Soriano, P. (1997). The PDGF alpha receptor is required for neural crest celldevelopment and for normal patterning of the somites. Development 124,2691-2700.

Stockard, C. (1910). The influence of alcohol and other anaesthetics onembryonic development. Am. J. Anat. 10, 369-392.

Streissguth, A. P. and Dehaene, P. (1993). Fetal alcohol syndrome in twins ofalcoholic mothers: concordance of diagnosis and IQ. Am. J. Med. Genet. 47,857-861.

Sulik, K. K. (2005). Genesis of alcohol-induced craniofacial dysmorphism. Exp.Biol. Med. (Maywood) 230, 366-375.

Sulik, K. K., Johnston, M. C., Daft, P. A., Russell, W. E. and Dehart, D. B. (1986).Fetal alcohol syndrome and DiGeorge anomaly: critical ethanol exposureperiods for craniofacial malformations as illustrated in an animal model. Am. J.Med. Genet. Suppl. 2, 97-112.

Swartz, M. E., Sheehan-Rooney, K., Dixon, M. J. and Eberhart, J. K. (2011).Examination of a palatogenic gene program in zebrafish. Dev. Dyn. 240, 2204-2220.

Swayze, V. W., II, Johnson, V. P., Hanson, J. W., Piven, J., Sato, Y., Giedd, J. N.,Mosnik, D. and Andreasen, N. C. (1997). Magnetic resonance imaging ofbrain anomalies in fetal alcohol syndrome. Pediatrics 99, 232-240.

Tallquist, M. and Kazlauskas, A. (2004). PDGF signaling in cells and mice.Cytokine Growth Factor Rev. 15, 205-213.

Tallquist, M. D. and Soriano, P. (2003). Cell autonomous requirement forPDGFRalpha in populations of cranial and cardiac neural crest cells.Development 130, 507-518.

Theveneau, E. and Mayor, R. (2012). Neural crest delamination and migration:from epithelium-to-mesenchyme transition to collective cell migration. Dev.Biol. 366, 34-54.

Thomas, J. D., Melcer, T., Weinert, S. and Riley, E. P. (1998). Neonatal alcoholexposure produces hyperactivity in high-alcohol-sensitive but not in low-alcohol-sensitive rats. Alcohol 16, 237-242.

Tittle, R. K., Sze, R., Ng, A., Nuckels, R. J., Swartz, M. E., Anderson, R. M.,Bosch, J., Stainier, D. Y., Eberhart, J. K. and Gross, J. M. (2011). Uhrf1 andDnmt1 are required for development and maintenance of the zebrafish lens.Dev. Biol. 350, 50-63.

Van Stry, M., McLaughlin, K. A., Ataliotis, P. and Symes, K. (2004). Themitochondrial-apoptotic pathway is triggered in Xenopus mesoderm cellsdeprived of PDGF receptor signaling during gastrulation. Dev. Biol. 268, 232-242.

Van Stry, M., Kazlauskas, A., Schreiber, S. L. and Symes, K. (2005). Distincteffectors of platelet-derived growth factor receptor-alpha signaling arerequired for cell survival during embryogenesis. Proc. Natl. Acad. Sci. USA 102,8233-8238.

Varga, Z. M., Amores, A., Lewis, K. E., Yan, Y. L., Postlethwait, J. H., Eisen, J. S.and Westerfield, M. (2001). Zebrafish smoothened functions in ventral neuraltube specification and axon tract formation. Development 128, 3497-3509.

Vary, T. C., Deiter, G. and Lantry, R. (2008). Chronic alcohol feeding impairsmTOR(Ser 2448) phosphorylation in rat hearts. Alcohol. Clin. Exp. Res. 32, 43-51.

Viljoen, D. L., Carr, L. G., Foroud, T. M., Brooke, L., Ramsay, M. and Li, T. K.(2001). Alcohol dehydrogenase-2*2 allele is associated with decreasedprevalence of fetal alcohol syndrome in the mixed-ancestry population of theWestern Cape Province, South Africa. Alcohol. Clin. Exp. Res. 25, 1719-1722.

Walker, M. B. and Kimmel, C. B. (2007). A two-color acid-free cartilage andbone stain for zebrafish larvae. Biotech. Histochem. 82, 23-28.

Warren, K. R. and Li, T. K. (2005). Genetic polymorphisms: impact on the risk offetal alcohol spectrum disorders. Birth Defects Res. A Clin. Mol. Teratol. 73, 195-203.

Westerfield, M. (1993). The Zebrafish Book: A Guide For the Laboratory Use ofZebrafish (Brachydanio Rerio). Eugene, OR: University of Oregon Press. D

EVELO

PMENT

3265RESEARCH ARTICLEProtective role for Pdgfra

Wu, E., Palmer, N., Tian, Z., Moseman, A. P., Galdzicki, M., Wang, X., Berger,B., Zhang, H. and Kohane, I. S. (2008). Comprehensive dissection of PDGF-PDGFR signaling pathways in PDGFR genetically defined cells. PLoS ONE 3,e3794.

Xiong, W., Cheng, B. H., Jia, S. B. and Tang, L. S. (2010). Involvement of thePI3K/Akt signaling pathway in platelet-derived growth factor-inducedmigration of human lens epithelial cells. Curr. Eye Res. 35, 389-401.

Xu, J., Yeon, J. E., Chang, H., Tison, G., Chen, G. J., Wands, J. and de laMonte, S. (2003). Ethanol impairs insulin-stimulated neuronal survival in

the developing brain: role of PTEN phosphatase. J. Biol. Chem. 278, 26929-26937.

Yoshida, T., Vivatbutsiri, P., Morriss-Kay, G., Saga, Y. and Iseki, S. (2008). Celllineage in mammalian craniofacial mesenchyme. Mech. Dev. 125, 797-808.

Zhang, C., Ojiaku, P. and Cole, G. J. (2013). Forebrain and hindbraindevelopment in zebrafish is sensitive to ethanol exposure involving agrin, Fgf,and sonic hedgehog function. Birth Defects Res. A Clin. Mol. Teratol. 97, 8-27.

Zhou, H. and Huang, S. (2010). The complexes of mammalian target ofrapamycin. Curr. Protein Pept. Sci. 11, 409-424.

DEVELO

PMENT